BeSweet v1.4b8 & Dolby Surround 2

[tiki4]

Hi there,

I downloaded the Downmix Plugin 0.1 from DSPguru's plugin site. Well, I read the included readme and then I searched the forums for some more specific information about that plugin. I didn't find information so far, so may I ask form some information about the plugin and how to use it? The readme mentions DPL2 downmix, so has anyone used it before and can share his experiences?

Any reply is welcome.

tiki4

[tiki4]

O.K.,

I found the answer to my question over at hydrogenaudio.org. Seems to be just for testing (for the programmers) and not for general use:(

tiki4

[DSPguru]

you don't have to be a developer in order to test :D.
just take some tracks and encode them to stereo waves, using the plugin. then test it with your surround system :).

[user]

Yeah,

I confirm that surround (rear) left and surround (rear) right channels are separated.

It seems after some testing, that plugin works fine.

The separation is probably the same as got by DS2 of headac3he0.23.


umm, why have I not been able to hear it some time earlier ?!
Perhaps some errors in my command line....


The working line you can find in logfile:

One example:

d:\own files\movies\programs\BeSweet ac3 to mp2\

BeSweet.exe -core( -input d:\own files\My Music\AC3\tests 5.1\1 channel ac3\sl.ac3 -output d:\own files\My Music\AC3\tests 5.1\sl.wav -2ch -logfile D:\own files\movies\programs\BeSweet ac3 to mp2\BeSweet.log ) -plugin( -name BS_Downmix.dll -func NoLFE -6ch -azid( -s surround -L 0db -S 0db --maximize ) -ssrc( --rate 44100 --twopass )




BeSweet v1.4b1 by DSPguru.
--------------------------
Using azid.dll v1.8 (b825) by Midas (midas@egon.gyaloglo.hu).
Using ssrc.dll v1.28 by Naoki Shibata & Dg (shibatch@geocities.co.jp).
Using BS_Downmix.dll v0.1 by DSPguru (http://DSPguru.doom9.org).

Logging start : 06/16/02 , 01:28:02.

d:\Eigene Dateien\Filme\DVD Video Audio Programme\BeSweet ac3 to mp2\BeSweet.exe -core( -input d:\Eigene Dateien\My Music\AC3\tests 5.1\dspguru 1 kanal ac3s\sl.ac3 -output d:\Eigene Dateien\My Music\AC3\tests 5.1\sl.wav -2ch -logfile D:\Eigene Dateien\Filme\DVD Video Audio Programme\BeSweet ac3 to mp2\BeSweet.log ) -plugin( -name BS_Downmix.dll -func NoLFE -6ch -azid( -s surround -L 0db -S 0db --maximize ) -ssrc( --rate 44100 --twopass )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : d:\Eigene Dateien\My Music\AC3\tests 5.1\dspguru 1 kanal ac3s\sl.ac3
[00:00:00:000] | Output: d:\Eigene Dateien\My Music\AC3\tests 5.1\sl.wav
[00:00:00:000] | Floating-Point Process: No
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby surround compatible
[00:00:00:000] | Total Gain: 0.0dB, Compression: None
[00:00:00:000] | LFE levels: To LR 0.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: No
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] +-------- SSRC -------
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] | Dest. Sample-Rate: 44.1KHz
[00:00:00:000] | Attenuation : 0.0db
[00:00:00:000] +---------------------
[00:00:04:352] Conversion Completed !
[00:00:04:352] Actual Avg. Bitrate : 1420kbps
[00:00:03:000] <-- Transcoding Duration
Logging ends : 06/16/02 , 01:28:05.

[DSPguru]

okay, BeSweet v1.4b7 (http://besweet.notrace.dk) now supports the downmix plugin in azid's section.

example :BeSweet v1.4b7 by DSPguru.
--------------------------
Using azid.dll v1.8 (b825) by Midas (midas@egon.gyaloglo.hu).
Using lame_enc.dll v1.28 (4/6/2002), Engine 3.92 <http://www.mp3dev.org/>.
Using BS_Downmix.dll v0.1 by DSPguru (http://DSPguru.doom9.org).

Logging start : 06/16/02 , 00:29:07.

E:\AC3TOMP3\BESWEET.EXE -core( -input e:\ac3tomp3\inputs\audio.ac3 -output test.mp3 -logfile test.txt ) -azid( -s surround2 )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : e:\ac3tomp3\inputs\audio.ac3
[00:00:00:000] | Output: test.mp3
[00:00:00:000] | Floating-Point Process: Yes
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby surround compatible
[00:00:00:000] | Total Gain: 0.0dB, Compression: None
[00:00:00:000] | LFE levels: To LR 0.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: No
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] +-------- LAME -------
[00:00:00:000] | Bitrate method : CBR
[00:00:00:000] | MP3 bitrate : 128
[00:00:00:000] | Channels Mode : Joint Stereo
[00:00:00:000] | Error Protection: No
[00:00:00:000] +---------------------
[00:00:21:984] Conversion Completed !
[00:00:21:984] Actual Avg. Bitrate : 128kbps
[00:00:04:000] <-- Transcoding Duration
Logging ends : 06/16/02 , 00:29:11.

thank you for confirming, user :D !

[ChAoS Overlord]

I have two questions about this.

a) How can I force decoding of this wav/mp3/ogg/... by the intervideo audio codec (because that one supports DPL2, doesn't it?).
b) what mp3 / ogg bitrate/quality setting should I use to preserve the DPL2 info?

[user]

I think 128 kbit/s mp3 should do it. The more, the better the sound quality in mp3. Use --alt-presets.

Every decoder is fine. The information of DS and DS2 you will keep even in analogue cables.


To get really DS2 or DS you need of course an amp/receiver with possibility of DS or DS2, like: Logic7 (Lexicon, Harman Kardon), DPL2 (Dolby), Circle Surround (Kenwood).
At PC: The new versions of WinDVD and PowerDVD are able of decoding DS2 information, too.

[ChAoS Overlord]

Originally posted by user
At PC: The new versions of WinDVD and PowerDVD are able of decoding DS2 information, too.

I know, that's why I want a way to force the decoding of the stream by those decoders. I have a SB Audigy too.

[user]

You have to look in menus of these programs. I have read, that the switch for DS2 decoding is a little bit hided in menus, but it should be able to find.

[ZeB]

What do you guys mean with Dolby Surround 2? Is it Dolby Pro Logic 2?

[user]

DS2 is the way (found by Dark Avenger) to encode 5.1 channels to stereo- and Dolby(Pro Logic)Surround(old)- compatible- waves, so that rear channels get some separation if this stereo-DS2-wave is decoded by Logic7/DPL2 (Dolby ProLogic2)/Circle Surround (Kenwood).

[jimbokoz]

I've found something to consider - your new plugin appears to be working quite well.

I'm currently using PowerDVD v4 to decode Dolby Surround media for my 6-channel analog sound system (based on SB Audigy). I unfortunately, don't have a hardware Pro-Logic decoder. Unfortunately, PowerDVD's DPL2 decoder only seems to decode old Dolby Surround media according to original specs - I only get front left/right and mono rear.

I've just attempted to transcode an AC3 stream using your plugin and tested it using PowerDVD's DPL2 decoder. Amazingly, it made a great difference - the centre channel is now being processed, and their appears to be a reasonable amount of channel separation at the rear speakers.

Just a couple of questions if I may:

1. What does the encoder do with the LFE information - is it encoded with the L+R at -3db as for older encoders, or is the information discarded?

2. Is there something I could be overlooking to get this result?

Thanks DSPguru - you appear to be having success at encoding DPL2 compatible streams...

[ZeB]

Originally posted by user
DS2 is the way (found by Dark Avenger) to encode 5.1 channels to stereo- and Dolby(Pro Logic)Surround(old)- compatible- waves, so that rear channels get some separation if this stereo-DS2-wave is decoded by Logic7/DPL2 (Dolby ProLogic2)/Circle Surround (Kenwood).

Thanks, now i understand...

[DSPguru]

with current plugin, the LFE isn't downmixed at all.
later this week, i will probably intergrate the downmix routines inside BeSweet and you could choose the LFE downmix level.

[frank]

Where can I find some information about the DS2 downmix values in the matrix?
Dolby's secret?

[DSPguru]

the source code of my downmix plugin is open.
download it from my webpage (http://dspguru.notrace.dk/effects.html).

[frank]

Yes, I know.
But why the -17 dB to sl and the -20 dB to sr??
No symmetric levels?
I searched the Internet, but didn't found any discussion.

BTW:
-3 dB = 20 lg a2/a1 -> 0.7079, not 0.6436 (-3.83 dB)
All values differ at -0.83 dB. Has no big effect to the results, only level is 0.83 dB lower.

[jimbokoz]

when it comes to the questions of LFE encoding into the L+R channels...

If one were to select -3db LFE into LR channels in Azid.dll settings, shouldn't the DLL premix that data into LR before passing it to the DPL2 encoder? Would this bypass the need for downmixing by the plugin itself?

I don't know if I can make this assumption about your program, DSP, but I'm going strong testing it out. I'm itching to use its features for encoding....

[frank]

but not required!
Dolby Digital Professional Encoding Guidelines
Section Consumer Decoder Products

3.4 Compatibility
...Whenever downmixing takes place, the LFE signal is discarded. Essential LFE
program content must be included in the main Left and Right channels to ensure that
it will be heard by all listeners. The LFE channel is never required in a program, and
is not an option in mono or stereo modes.

3.5 LFE and Bass Management
...The LFE channel, if present in the encoded
audio program, is always omitted during playback from a two-channel product.andDolby Digital Professional Encoding Guidelines

4.5.3 LFE Channel
The LFE Channel parameter enables or disables the Low-Frequency Effects (LFE)
channel. Use of the LFE channel is optional with multichannel programs, but is not
available for mono, stereo, or surround-encoded programs.
Two-channel Dolby Digital products, or multichannel products operating in a two-
channel downmix mode, omit the LFE signal. Therefore, low-frequency content
essential to the program should never be mixed exclusively to the LFE channel. Refer
to Section 3.5, LFE and Bass Management, for more information on decoder handling
of LFE signals.

SUB out of DS2 is derived from the Lt Rt channels and not equal to LFE.

[DSPguru]

Originally posted by frank
Yes, I know.
But why the -17 dB to sl and the -20 dB to sr??
No symmetric levels?
I searched the Internet, but didn't found any discussion.this assymetirc valus is the main difference between dpl1 to dpl2.
it's for the l/r seperation.

BTW:
-3 dB = 20 lg a2/a1 -> 0.7079, not 0.6436 (-3.83 dB)
All values differ at -0.83 dB. Has no big effect to the results, only level is 0.83 dB lower. 20 lg a2/a2. sure. but what's the base of your log ;) ?

anyway, i scaled it because 0.1441+0.6436+ 0.1284+0.0909=1 :)

Originally posted by jimbokoz
when it comes to the questions of LFE encoding into the L+R channels...

If one were to select -3db LFE into LR channels in Azid.dll settings, shouldn't the DLL premix that data into LR before passing it to the DPL2 encoder? Would this bypass the need for downmixing by the plugin itself?sounds reasonable, you're probably right !!

I don't know if I can make this assumption about your program, DSP, but I'm going strong testing it out. I'm itching to use its features for encoding.... waiting for your feedbacks...


@frank
you're right, but there are still ppl who are interested in this feature.

[frank]

-0.83 dB:
Oh yes, I forgot the prevention of arithmetic overflow. :)
But the probability that all channels have max is 0.00...

I look at the attenuation/amplification of the downmix because I decode all my tracks with fixed compensating values like hardware does. All my disks, cuts and pastes have the same sound level! No overflow.
The time wasting level normalization is not nessecary and not professional. AC3 has calibrated levels.

LFE: Ey, I think you are a professional guru. This ppl don't know electronics. I also know a lot of ppl who want AC3 tracks on SVCD! But standard and hardware say NO... ;)

[DSPguru]

preventing overflow is so trivial, that dolby Almost disregards it :) :Prior to the scaling needed to prevent overflow, the 3/2 downmix equations for an
LtRt stereo signal are:
Lt = 1.0 * L + 0.707 * C - 0.707 * Ls - 0.707 * Rs ;
Rt = 1.0 * R + 0.707 * C + 0.707 * Ls + 0.707 * Rs ;
btw, in your case, you could simply use a static gain value of 0.83dB.

Cheers :D,
Dg.

[jimbokoz]

DSP Guru: I can confirm that Azid.dll appears to be mixing LFE data into the L and R channels before passing all six channels to your DPL2 downmixer.

I tested this with OggMachine 0.3, BeSweet 1.4b7 (listed as 1.4b6 in the log???), on a 10 min section of "Stargate (lots of bass). I played the resulting Ogg files in Power DVD 4 for testing.

I did 3 encodes. The first, has "LFE to LR" checked, and at -3db.
The second has "LFE to LR" unchecked.
The third I did querying a bug in the output of the previous encode, listing the LFE to LR as being performed at 0db. So, for this encode, I checked "LFE to LR" and set it to -infdb. I don't know if it did anything...!

Sorry about having SSRC enabled, forgot I had it checked. Shouldn't make a difference...

Here are the log files:

BeSweet v1.4b6 by DSPguru.
--------------------------
Using azid.dll v1.8 (b825) by Midas (midas@egon.gyaloglo.hu).
Using ssrc.dll v1.28 by Naoki Shibata & Dg (shibatch@geocities.co.jp).
Using libVorbis.dll by John33 (http://www.inf.ufpr.br/~rja00).

Logging start : 06/19/02 , 11:13:22.

C:\Program Files\DivX\BeSweet\BeSweet.exe -core( -input D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ac3 -output D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ogg -logfilea BeSweet.log ) -azid( -L -3db -n1 -g max -c normal -s surround2 ) -ssrc( --rate 44100 ) -split( -start 3600 -end 4200 ) -ogg( -q 1.000 )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ac3
[00:00:00:000] | Output: D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ogg
[00:00:00:000] | Floating-Point Process: Yes
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby surround compatible
[00:00:00:000] | Total Gain: 12.3dB, Compression: Normal
[00:00:00:000] | LFE levels: To LR -3.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: Yes
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] +-------- SSRC -------
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] | Dest. Sample-Rate: 44.1KHz
[00:00:00:000] | Attenuation : 0.0db
[00:00:00:000] +-------- OGG --------
[00:00:00:000] | VBR Quality : 1.000
[00:00:00:000] +---------------------
[00:09:59:968] Conversion Completed !
[00:09:59:968] Actual Avg. Bitrate : 401kbps
[00:03:35:000] <-- Transcoding Duration
Logging ends : 06/19/02 , 11:16:57.

File Size: 30,095,415 bytes

BeSweet v1.4b6 by DSPguru.
--------------------------
Using azid.dll v1.8 (b825) by Midas (midas@egon.gyaloglo.hu).
Using ssrc.dll v1.28 by Naoki Shibata & Dg (shibatch@geocities.co.jp).
Using libVorbis.dll by John33 (http://www.inf.ufpr.br/~rja00).

Logging start : 06/19/02 , 11:17:21.

C:\Program Files\DivX\BeSweet\BeSweet.exe -core( -input D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ac3 -output D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ogg -logfilea BeSweet.log ) -azid( -n1 -g max -c normal -s surround2 ) -ssrc( --rate 44100 ) -split( -start 3600 -end 4200 ) -ogg( -q 1.000 )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ac3
[00:00:00:000] | Output: D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ogg
[00:00:00:000] | Floating-Point Process: Yes
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby surround compatible
[00:00:00:000] | Total Gain: 12.1dB, Compression: Normal
[00:00:00:000] | LFE levels: To LR 0.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: Yes
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] +-------- SSRC -------
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] | Dest. Sample-Rate: 44.1KHz
[00:00:00:000] | Attenuation : 0.0db
[00:00:00:000] +-------- OGG --------
[00:00:00:000] | VBR Quality : 1.000
[00:00:00:000] +---------------------
[00:09:59:968] Conversion Completed !
[00:09:59:968] Actual Avg. Bitrate : 399kbps
[00:03:33:000] <-- Transcoding Duration
Logging ends : 06/19/02 , 11:20:54.

File Size: 29,984,837 bytes

BeSweet v1.4b6 by DSPguru.
--------------------------
Using azid.dll v1.8 (b825) by Midas (midas@egon.gyaloglo.hu).
Using ssrc.dll v1.28 by Naoki Shibata & Dg (shibatch@geocities.co.jp).
Using libVorbis.dll by John33 (http://www.inf.ufpr.br/~rja00).

Logging start : 06/19/02 , 11:21:20.

C:\Program Files\DivX\BeSweet\BeSweet.exe -core( -input D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ac3 -output D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ogg -logfilea BeSweet.log ) -azid( -L -infdb -n1 -g max -c normal -s surround2 ) -ssrc( --rate 44100 ) -split( -start 3600 -end 4200 ) -ogg( -q 1.000 )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ac3
[00:00:00:000] | Output: D:\Movies\Stargate\vts_01_([0x80]_Audio_English_AC3(6Ch)_48kHz___)_Delay_0ms.ogg
[00:00:00:000] | Floating-Point Process: Yes
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby surround compatible
[00:00:00:000] | Total Gain: 10.7dB, Compression: Normal
[00:00:00:000] | LFE levels: To LR -1.$dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: Yes
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] +-------- SSRC -------
[00:00:00:000] | Source Sample-Rate: 48.0KHz
[00:00:00:000] | Dest. Sample-Rate: 44.1KHz
[00:00:00:000] | Attenuation : 0.0db
[00:00:00:000] +-------- OGG --------
[00:00:00:000] | VBR Quality : 1.000
[00:00:00:000] +---------------------
[00:09:59:968] Conversion Completed !
[00:09:59:968] Actual Avg. Bitrate : 402kbps
[00:03:34:000] <-- Transcoding Duration
Logging ends : 06/19/02 , 11:24:54.

File Size: 30,183,815 bytes








As you can see, each setting had a different max gain detected, and a different bitrate and filesize as encoded by vorbis, indicating the actual audio data in each final result is slightly different.

Frank:

In deference to your apparent expertise in this area, when listening to the resulting files, I could not hear any difference in bass levels between them. Maybe I need a bigger sound system??? :-)

If there's a problem with the above methodology, or the conclusions I have drawn from it, please let me know.

[frank]

There is an issue in Azid. If LFE to LR is unchecked, then Azid sets his standard level = 0 dB rather than 0!! Equals 1.0!! You have to set -L 0 if you don't want the LFE to LR.
In Azid's downmix matrix you can see the real working components.
(But with 1.4b7 + BS_downmix.dll and "-M1 -s surround2" you'll get no display)

[yaydo]

Hi,

I am new to DVD2SVCD and BeSweet. I want to get Dolby Digital audio in my SVCDs and I was reading the postings in this thread but I can not figure out how to set the command line options you are talking about in DVD2SVCD. Can you help? What shall I do to get BeSweet use the new DOwnmix plugin in DVD2SVCD GUI?

Thanks.
Yaydo

[DSPguru]

@jimbokoz
thank you for your inputs! i guess i completely forgot about the lfe2lr behavior :).

@frank
you are right. will fix this in beta8.
and sorry for being lazy here, but i'm not planning to print the downmix matrix. (don't forget you could modify anything in the downmix.dll source)

@yaydo
best thing would be if you had created a batchfile in the name BeSweet.bat that would look like something :BeSweet.exe %1 %2 %3 %4 %5 %6 %7 %8 %9 %10 %11 %12 %13 %14 %15 %16 %17 %18 %19 %20 -plugin( -name BS_Downmix.dll -func NoLFE -6ch )
make sure to use the latest beta of BeSweet and having the BS_Downmix.dll at the same folder.

[Beave]

I'm still only having a stereo system here, but I will upgrade to a 5.1 pretty soon (hopefully, since I've got an Audigy in my system ideling).
Is it save to start using DS2 encoding yet? As I understand the stereo signal won't be affected.
Will the commandline change towards the final release? I'm using quite some VB Scripts in my Send To folder, which I have to change individually. That's why I ask.

[Farok]

Hi !

this plugin is made for my Denon :)
But DspGuru, are you planning to enhance the speed of your plugin, because it is really slow on my Tbird 1.4Ghz (maybe 1sec to encode 2sec of sound :( )

[frank]

@DspGuru
Yes, if you convert AC3 3/2 to dpl2 2/0 there is basicly no need for matrix dump, because BM_downmix.dll does the work.
All Azid output channels (3/2) have to be at max, no Azid downmix needed.

But I'm wondering: At conversion to dpl2 I had to compensate the old Azid matrix attenuation of -11.7 dB!
May be the attenuation is active because Azid works on his standard mode 2/0?
I think this attenuation must not be in 3/2 decoding mode -> lost bits.

In the result I had to set +11.7 dB + 0.83 dB = 12.5 dB gain for compensation. With this value I achieved the original track level. (-27 dB normal)

Cheers
frank

[DSPguru]

BeSweet v1.4b8 now has internernal implementation of DS2.
haven't debugged it, though. so, please report any bugs.

the switch synatx is set to -azid( -s surround2 ) and won't change.

@frank
lost bits ? ain't we talking about the floating-point mode ?
i guessed i didn't figure your encoding method...

[frank]

Sorry I was confused, Azid works in fp mode:)
But this damned attenuation is annoying.
Hey my method: DVD2AVI -> BeSweet wav, fade-in/out - mp2 -> bbMPEG muxer -> TSCV 0.84 -> SVCD -> DVD + Pioneer VXS-859RDS (5x100W) ->Fun :D :D
But my receiver decodes only pro logic :(

Oh it starts to rain and thunder... must end now.

[DSPguru]

Originally posted by frank
Sorry I was confused, Azid works in fp mode:)
But this damned attenuation is annoying.
Hey my method: DVD2AVI -> BeSweet wav, fade-in/out - mp2 -> bbMPEG muxer -> TSCV 0.84 -> SVCD -> DVD + Pioneer VXS-859RDS (5x100W) ->Fun :D :D
But my receiver decodes only pro logic :("BeSweet wav" ?
btw, if you always use a static gain value and, why do you care about its value :) ?

Oh it starts to rain and thunder... must end now. fun fun fun !

[frank]

My computer survived the thunderstorm :)

With DVD2AVI I rip the AC3 track parts fitting one CD. No cut program needed. I decode to WAVs because I want to edit the parts -> fade-in/out.
Azid prevents in the downmix modes arithmetic overrun. Result in different attenuations that need different gain to compensate!
Ex: 3/2 gain=0 dB, 2/0 gain=11.7 dB

Latest test with b8:
The static gain to keep the -27 dB dialogue level in ds2 is about 10.5 dB. I don't know why, because Azid decodes every channel in 3/2 mode WITHOUT any attenuation.
I compared the avr rms power with old ds downmix mode (+11.7 dB gain).

[DSPguru]

does the DS2 inner integration seem ok ? did you notice any speed improvement ?

about fade-in,
you could write a BeSweet plug-in for that. there's a fadein source-code example in my page.

[frank]

Seems to be ok, but will test the speed when I'll back at home.

DS2 downmix:
I have the feeling that the centre channel ist too much to the fore.
(downmix +13 dB above l,r ? )
At the moment I can only hear in ds pro logic, but Dolby swears that ds2 is compatible and sounds well on pro logic devices.

[DSPguru]

yea, that's weird.
but ppl reports success :).

[FiW]

delay in besweet is great...
fade/in-out in besweet is great...
hmm... and needed LENGTH! for crop credits

---
bad english... ;-)

[trg100]

I have been experimenting with this DS2 downmix concept. I have a four-way SBLive and speakers and would like to put quad channel audio on DVD rips. I'm still waiting for the appropriate multichannel coupling option in Ogg Vorbis but if DS2 can approximate DD 5.1 with a 2ch file then it's an interesting option.

I made myself a 5.1 channel AC3 test file in SoftEncode then used the -surround2 switch in BeSweet (1.4b8) to decode and downmix to a DS2 wav file.

I used PowerDVD 4.0 in 4-speaker mode with DPL2 enabled to play back the files. While the channel placement in the AC3 is perfect it's all wrong on the DS2 wav. For example, the right channel is missing and surround left comes from the right speaker.

Any ideas what might have gone wrong? I can be more specific if anyone is interested.

I wanted to test the downmix in HeadAC3che too but it says my AC3 is corrupt(?!)

[DSPguru]

logfile ?

[trg100]

BeSweet log file looks like this. I hope this is what you mean. If there's something I should do to make it more informative let me know (I'm potentially out of my depth!). Cheers.

BeSweet v1.4b8 by DSPguru.
--------------------------
Using azid.dll v1.8 (b825) by Midas (midas@egon.gyaloglo.hu).

Logging start : 06/21/02 , 19:58:11.

C:\Program Files\BeSweet\BeSweet.exe -core( -input f:\Audio\Sound\Quad tests\5point1.ac3 -output f:\Audio\Sound\Quad tests\5point1_ds2.wav -2ch -logfile C:\Program Files\BeSweet\BeSweet.log ) -azid( -z1 -b1 -M1 -s surround2 -L -3db )

[00:00:00:000] +------- BeSweet -----
[00:00:00:000] | Input : f:\Audio\Sound\Quad tests\5point1.ac3
[00:00:00:000] | Output: f:\Audio\Sound\Quad tests\5point1_ds2.wav
[00:00:00:000] | Floating-Point Process: No
[00:00:00:000] +-------- AZID -------
[00:00:00:000] | Output Stereo mode: Dolby Surround 2 compatible
[00:00:00:000] | Total Gain: 0.0dB, Compression: None
[00:00:00:000] | LFE levels: To LR -3.0dB, To LFE 0.0dB
[00:00:00:000] | Center mix level: BSI
[00:00:00:000] | Surround mix level: BSI
[00:00:00:000] | Dialog normalization: No
[00:00:00:000] | Rear channels filtering: No
[00:00:00:000] | Source Sample-Rate: 44.1KHz
[00:00:00:000] +---------------------
[00:00:06:965] Conversion Completed !
[00:00:06:965] Actual Avg. Bitrate : 1411kbps
[00:00:01:000] <-- Transcoding Duration
Logging ends : 06/21/02 , 19:58:12.

[frank]

May be you did the channel order mismatch in Softencode.
Proof it, or take a ripped AC-3 stream.

In Winamp, and on my SVCDs the channel order of surround2 was ok.

Oh, correction: tested in B7.

[frank]

And here some benchmarks:

Test environment:
P4 2000, Win XP, BeSweet + SSRC, surround2,
File AC-3 5.1, 49:24


BS 1.4b7: WAV 400s MP2 508s ----- 100%
BS 1.4b8: WAV 433s MP2 515s --> +8.3% / +1.4%
BS 1.4b9: WAV 426s MP2 528s --> +6.5% / +3.9%
BS 1.4b11 WAV 426s MP2 529s --> +6.5% / +4.1%


BTW, conversion to WAVs always showes floating point off.:confused:

[trg100]

Thanks for the responses

I checked the channel order was right in softencode and tried a ripped AC3 track and the results are the same. PowerDVD's 4 speaker mode works great with the 5.1 AC3 but with the DS2 stereo wav and DPL2 the channel mapping is wrong.

DD 5.1 AC3 mapped to 4 channel SBLive (FrontL, FR, RearL, RR)
L -> FL
C -> FL (much louder than anything else)
R -> it's there but it's very quiet and muffled and swishes about the channels
SL -> FR
SR -> RL
LFE -> FL (though mostly filtered to woofer on 4.'1' speaker system)

If use 2 speaker mode then all then channels are where you'd expect on a stereo axis (if you know what I mean).

It's possible I am missing something but it's not obvious to me.

[trg100]

I just got the DS2 downmix working in headac3he and results are the same.

Sorry, I guess this is not a particularly mainstream problem! I suppose trying the intervideo filter might be interesting.

[DSPguru]

Originally posted by frank
BeeSweet 1.4B7: WAV 400s MP2 508s
BeeSweet 1.4B8: WAV 433s MP2 515s -- +8.3% / +1.4%

:( What's happen Guru, B8 slowes down?could be..
i had to compensate ugly behavior of azid when doing 2pass.
azid used to saturate some of the signals on the second pass.
i prefer quality over speed, that's for sure.
will think of a better solution..


BTW, conversion to WAVs always showes floating point off.:confused: that's true. the wav files are not fp wavs.

[frank]

i prefer quality over speed, that's for sure.I agree.:cool:

I had the same problem with clipping. I had to split my gain in two parts: Azid +6 dB and -ota +4.5 dB.

[DSPguru]

since you don't do two-pass, you might wanna keep using the external plugin.